The collapse of Sam Bankman-Fried’s crypto empire, FTX, has been stunning. Just a few weeks ago, the firm was one of the top players in the industry. Bankman-Fried was being compared to JP Morgan and Warren Buffett. He had the financial backing of some of the world’s top firms, like Sequoia and BlackRock. He ranked among the wealthiest persons in the world. But in a couple days, FTX would come undone like a house of cards. According to a report from the Financial Times, the liabilities were about $9 billion and liquid assets at less than $1 billion. Bankruptcy was the only option.
The consequences of this crash are far from clear. But it seems likely that governments will push for regulation of the industry, such as by mandating better controls and protections for customer funds. In the meantime, venture capital funding will dry up. But there is something else that may not be so obvious. The crypto industry – as well as related categories like fintech and DeFi – will need to focus much more on customer service and back-office operations. And this will mean adopting technologies like conversational AI. If not, it could be very difficult for the industry to make a meaningful comeback.
When it comes to traditional financial services, customer support and back-office technology has been essential for growth. An example of this is the “Paperwork Crisis” of the late 1960s. With the bull market, the trading volume surged. But the manual systems – which relied on messengers to deliver stock certificates — buckled. The situation got so bad that the New York Stock Exchange limited trading to four days a week. On some days, the hours had to be cut back.
But these actions were woefully inadequate. There were various firms that had to close down because of failed trades.
To deal with this, the NYSE and Wall Street firms invested heavily in computers and software. There was also the creation of the Depository Trust Corporation (DTC), which provided a centralized system to settle trades.
Of course, the investments in automation would only continue. If anything, Wall Street has become one of the most technologically savvy industries.
Now, as for the crypto industry, there is a similar crisis for support and the back office. Because of the wide swings in activity, it can be extremely difficult to scale up and down. This is especially the case with traditional customer support ticketing systems.
Consider a comprehensive report from the Consumer Financial Protection Bureau (CFPB), which involved over 8,300 complaints from crypto customers during the past two years. It notes: “Poor customer service is a recurring issue in complaints about crypto assets. Consumers described hard to reach, non-responsive, or non-existent customer service. Some consumers reported being unable to get their problem resolved because they cannot reach a human at the company, and that the failure to provide timely customer service exposed them to unnecessary risks.”
What to do? Well, interestingly enough, conversational AI can be a big help. This is a category that has benefited from large amounts of investments — from megatech companies to startups. The result is that the technology is quite sophisticated. The pandemic also helped accelerate the progress as many companies had little choice but to automate.
Customers are also becoming accustomed to using Conversational AI. As an indication of this, Frontier Airlines recently dropped phone support. Instead, the first line of customer support is a chatbot.
True, as with any technology, Conversational AI will not be a silver bullet either. The technology cannot completely automate customer service. A conversational AI project will usually need some bespoke customization, such as with designing the taxonomy. A general purpose one is typically not sufficient enough.
Next, it can take a few months for the learnings of the conversational AI to make progress. There also needs to be ongoing monitoring of the models, so as to deal with model drift.
Yet conversational AI – when implemented properly and done with sufficient training – can bolster the human reps. To see how, take a look at Carta. The company operates a platform that allows over 2.1 million customers to manage their stock holdings. The company is part of the Forbes World’s Best Cloud Companies, Fast Company’s Most Innovative list, and Inc.’s Fastest-Growing Private Companies.
To manage the growth and keep customers satisfied, Carta has invested in a conversational AI system called AISERA (the company recently raised $90 million). This has involved a rethinking of the processes. For example, to use the conversational AI effectively, Carta has identified areas of low-hanging fruit for automation. The result is that a chatbot can handle a customer interaction in a couple minutes versus 20 minutes when using traditional approaches.
By leveraging conversational AI, Carta has been able to scale customer service while improving satisfaction scores. In fact, the self-service rate went from 46% to 74%. The has resulted in major savings, at about $1 million. But more importantly, Carta has been able to handle the uneven swings in customer inquiries.
Granted, for the crypto industry, there are many challenges to tackle. But to rebuild trust, there needs to be more focus on un-sexy areas like customer service and back-office operations. If not, there could be a prolonged winter for the industry.